How we got this number
Each generated token reads the active part of the model from memory once: 2.4 GB. The Mac mini M4 (16GB) moves 120 GB/s, so the ceiling is 50 tokens/sec. Real machines reach a fraction of that ceiling; we use 80%, calibrated against a measured Mac run. Send us your real numbers and we'll replace the estimate.
Faster machines for gpt-oss 20B
Other models on the Mac mini M4 (16GB)
| lfm 2.5 8B (A1B) | Fast | 78 tok/s |
| Qwen2.5-VL 3B | Fast | 30 tok/s |
| Gemma 3 4B | Fast | 29 tok/s |
| Qwen 3-VL 4B | Fast | 29 tok/s |
| translateGemma 4B | Fast | 29 tok/s |
| Qwen3.5 4B | Fast | 28 tok/s |